50 fixes a really obvious error in sample-level surplus result processing that should have been caught by automated testing (in a compiled language the package wouldn't even have built). We need better test coverage for these more niche result specs.
50 fixes a really obvious error in sample-level surplus result processing that should have been caught by automated testing (in a compiled language the package wouldn't even have built). We need better test coverage for these more niche result specs.